The successful delivery of thermal therapy in human patients using radiofrequency waves is limited by their complex interactions with different tissue types. These interactions make it difficult to focus heating in the tumor and avoid excessive temperatures in surrounding normal tissues. The hypothesis driving this project is that a novel approach combining finite element (FE) modeling and magnetic resonance thermal imaging (MRTI) with feedback guidance can be used to rapidly optimize patient heating in the MR unit for breast, extremity and pelvic sites. The specific aims supporting this hypothesis are: (1) Development and implementation of FE modeling and MRTI-guided thermal prediction in phantom verification experiments. This specific aim will create the individual tools (FE modeling, MRTI-guided prediction) used in this project. (2) Combination of the FE model and MRTI-guided prediction into a single rapid thermal model (RTM), and validation of the RTM in phantom experiments. This specific aim will extract useful information from the FE model to allow the heated body to be characterized by a simpler, reduced-order representation. This reduced-order representation will be used with MRTI to create the RTM. (3) Implement the RTM in combination with feedback correction in 30 patient treatments. Feedback correction will be used to minimize the difference between the optimal RTM-predicted and MR-measured temperature distributions until it is at an acceptable minimum. This project will interface closely with projects 1, 3, 5, and Cores B, D. Project 1 and Core D will supply the MR thermal imaging technology developments and heating equipment development, respectively. Project 5 will supply the human patients for Specific Aim 3, and Core B will evaluate the effectiveness of the RTM approach. Project 3 will seek to optimize therapeutic efficacy of thermolabile drug containing liposomes that are dependent on the spatial temperature distribution. This project will facilitate the challenging objective of optimizing thermal therapy in human patients.